Resource Allocation for Multimedia Services Over Cloud Computing
Total Page：16
File Type：pdf, Size：1020Kb
Resource Allocation for Multimedia Services over Cloud Computing by Xiaoming Nan M. Sc., Beijing University of Posts and Telecommunications, Beijing, China, 2010 B. Sc., Xidian University, Xi’an, China, 2007 A Dissertation presented to Ryerson University in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Program of Electrical and Computer Engineering Toronto, Ontario, Canada, 2015 c Xiaoming Nan 2015 I hereby declare that I am the sole author of this dissertation. This is a true copy of the dissertation, including any required final revisions, as accepted by my examiners. I authorize Ryerson University to lend this dissertation to other institutions or individuals for the purpose of scholarly research. I further authorize Ryerson University to reproduce this dissertation by photocopying or by other means, in total or in part, at the request of other institutions or individuals for the purpose of scholarly research. I understand that my dissertation may be made electronically available to the public. ii Resource Allocation for Multimedia Services over Cloud Computing Doctor of Philosophy 2015 Xiaoming Nan Electrical and Computer Engineering Ryerson University Abstract Cloud-based multimedia application has emerged as a popular service, delivering on- demand media computing and storage to millions of users. Though widely deployed, the quality of service (QoS) in current cloud-based multimedia service is not satisfying, due to the varying user demands and strict response time requirements. This thesis investigates resource allocation approaches to improve QoS for cloud-based multimedia services. A service model is desired to quantify the user demands and resource allocation. To meet this need, we propose a queueing model to characterize the cloud service process, based on which we investigate the response time minimization problem and the resource cost minimization problem in single-service scenario, multi-service scenario, and priority- service scenario, respectively. Dynamic workload causes the unbalanced resource utilization and local congestion in multimedia cloud. To address this issue, we propose a two-time-scale resource configura- tion (TRC) scheme to dynamically allocate virtual machines (VMs) to adapt to varying iii workload. Based on the TRC scheme, we solve the optimal VM configuration problems to minimize the resource cost or minimize the average response time for the single-site cloud scenario and the multi-site cloud scenario, respectively. We propose optimal workload scheduling schemes at user level and task level, respec- tively. At user level, we optimize the workload assignment to minimize the response time or minimize the resource cost. At task level, we introduce a directed acyclic graph to model the precedence constraints among tasks, and then solve the execution time minimization problems for sequential structure, parallel structure, and mixed structure, respectively. Cloud gaming is an emerging interactive multimedia service. However, current cloud gaming suffers from a high bandwidth consumption and a large response delay. We propose a hybrid streaming framework to provide a high quality cloud gaming experience. We solve the delay-rate-distortion (d-R-D) optimization problem to minimize the overall distortion under the bandwidth and response delay constraints. iv Acknowledgements I would like to first express my sincere gratitude to my supervisor, Dr. Ling Guan, for his invaluable guidance and support throughout my Ph.D. study. I deeply appreciate his insightful discussions, precious suggestions, and strict training in all aspects of my academic development, which will be my life long treasure. I am very grateful to my co- supervisor, Dr. Yifeng He, for his rigorous attitude toward research and great attentions on details. Dr. He has taught me various research approaches and given me insightful comments. I feel very fortunate to have Dr. Guan and Dr. He as my Ph.D. supervisors. Their dedication, diligence, and enthusiasm will continue motivating me in my future career. I would also like to thank Dr. Baining Guo, Dr. Yan Lu, and Dr. Xun Guo, for their insightful ideas and discussions on cloud gaming project during my internship at Microsoft Research Asia. I have benefited immensely from their mentorship. It is a blessing for me to join Ryerson Multimedia Research Laboratory (RML). I have gained both knowledge and friendship during my study at RML. I am grateful to all my colleagues at RML for their valuable support and help. Last but not least, I would like to thank my parents for their unconditional support and encouragement. I am especially grateful to my beloved wife, Yi Liu, and my daughter, Stephanie Nan. Without their relentless love and support, I would never have been able to finish my thesis. v Contents Declaration ..................................... ii Abstract ....................................... iii Acknowledgements ................................. v List of Tables .................................... x List of Figures ................................... xi List of Appendices ................................. xv 1 Introduction 1 1.1 Motivation................................... 1 1.2 Objective ................................... 3 1.3 Challenges in Multimedia Cloud Computing . 4 1.4 MainContributions.............................. 6 1.5 ThesisOrganization.............................. 8 2 Background 11 2.1 CloudComputingOverview ......................... 11 2.1.1 Cloud Computing Definition . 11 2.1.2 ServiceModels ............................ 13 2.1.3 CloudComputingIssues . .. .. 14 2.1.4 Multimedia Cloud Computing . 15 2.2 Resource Allocation and Workload Scheduling . 17 2.2.1 ResourceAllocation.......................... 17 2.2.2 Workload Scheduling . 18 2.3 CloudGaming................................. 19 2.4 Queueing Theory in Resource Allocation . 21 vi 2.5 ChapterSummary .............................. 22 3 Queueing Model based Resource Optimization for Multimedia Cloud 23 3.1 Introduction.................................. 23 3.2 SystemModels ................................ 25 3.2.1 DataCenterArchitecture . 25 3.2.2 QueueingModel............................ 27 3.2.3 CostModel .............................. 28 3.3 Queueing Model based Resource Optimization . 28 3.3.1 Single-service Scenario . 29 3.3.2 Multi-service Scenario . 36 3.3.3 Priority-service Scenario . 40 3.4 Simulations .................................. 43 3.4.1 Simulations in Single-service Scenario . 43 3.4.2 Simulations in Multi-service Scenario . 45 3.4.3 Simulations in Priority-service Scenario . 47 3.5 ChapterSummary .............................. 49 4 Dynamic Resource Configuration for Cloud based Multimedia Services 50 4.1 Introduction.................................. 50 4.2 SystemModels ................................ 52 4.2.1 Multimedia Cloud Architecture . 52 4.2.2 VM Pricing Plan . 53 4.2.3 Two-time-scale Resource Configuration (TRC) Scheme . 54 4.2.4 WorkloadPredictionModel . 54 4.3 Dynamic Resource Configuration for Single-site Cloud . .. 55 4.3.1 Resource Allocation for Single-site Cloud . 55 4.3.2 Resource Reconfiguration for Single-site Cloud . 58 4.4 Dynamic Resource Configuration for Multi-site Cloud . 61 4.4.1 Resource Allocation for Multi-site Cloud . 61 4.4.2 Resource Reconfiguration for Multi-site Cloud . 62 4.5 PerformanceEvaluation ........................... 65 4.5.1 Simulations for Single-site Cloud . 66 vii 4.5.2 Simulations for Multi-site Cloud . 69 4.6 ChapterSummary .............................. 71 5 Optimization of Workload Scheduling over Multimedia Cloud 73 5.1 Introduction.................................. 73 5.2 Optimization of User Level Workload Scheduling . 75 5.2.1 User Level Workload Scheduling Model . 75 5.2.2 Response Time Minimization Problem . 76 5.2.3 Resource Cost Minimization Problem . 77 5.2.4 Simulations . 78 5.3 Optimization of Task Level Workload Scheduling . 81 5.3.1 Task Level Workload Scheduling Model . 81 5.3.2 Optimal Task Level Scheduling for Sequential Structure . .. 82 5.3.3 Optimal Task Level Scheduling for Parallel Structure . 83 5.3.4 Optimal Task Level Scheduling for Mixed Structure . 84 5.3.5 Heuristic for Optimal Task Level Scheduling . 84 5.3.6 Simulations . 85 5.4 ChapterSummary .............................. 88 6 Delay-rate-distortion Optimization for Cloud Gaming with Hybrid Stream- ing 89 6.1 Introduction.................................. 90 6.2 Hybrid Streaming Framework for Cloud Gaming . 92 6.3 Optimal Rate Allocation for Hybrid Streaming Framework . 96 6.3.1 Problem Formulation . 96 6.3.2 GamingProcess............................ 97 6.3.3 Rate-Distortion (R-D) Analysis . 100 6.3.4 ResponseDelayAnalysis . 102 6.3.5 Rate Allocation Algorithm . 105 6.4 PerformanceEvaluation . .. .. 110 6.4.1 ExperimentSetup........................... 110 6.4.2 Comparison with Video Streaming Approach . 111 6.4.3 Comparison with Graphics Streaming Approach . 116 viii 6.4.4 Evaluation of Rate Allocation Algorithm . 119 6.5 ChapterSummary .............................. 121 7 Conclusions and Future Work 126 7.1 Conclusions .................................. 126 7.2 FutureWork.................................. 128 7.2.1 Big Data on Multimedia Cloud . 128 7.2.2 Collaborative Media Computing and Rendering . 129 7.2.3 Joint Resource Allocation of Cloud and Network . 129 7.2.4 Optimization of Internal Traffic Management in Cloud . 129 7.2.5 Mobile multimedia applications . 130 References 151 ix List of Tables 1.1 Summaryofoptimizationproblems .